Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Improving Semantic Embedding Consistency by Metric Learning for Zero-Shot Classiffication
Maxime Bucher, Stéphane Herbin, Frédéric Jurie
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add the learned Mahalanobis metric (instead of Euclidean distance, W_A = identity) (ablation) | zero-shot metric learning model (VGG) · aPascal&aYahoo ZSL | accuracy: 47.25 -> 53.15 | Improved |
| add the learned Mahalanobis metric (instead of Euclidean distance, W_A = identity) (ablation) | zero-shot metric learning model (VGG) · AwA ZSL | accuracy: 73.81 -> 77.32 | Improved |
| add the learned Mahalanobis metric (instead of Euclidean distance, W_A = identity) (ablation) | zero-shot metric learning model (VGG) · CUB-200-2011 ZSL | accuracy: 33.87 -> 43.29 | Improved |
| add the learned Mahalanobis metric (instead of Euclidean distance, W_A = identity) (ablation) | zero-shot metric learning model (VGG) · SUN Attributes ZSL | accuracy: 74.91 -> 84.41 | Improved |
| add the attribute prediction term to the criterion (ablation) | zero-shot metric learning model (VGG) · aPascal&aYahoo ZSL | accuracy: 48.47 -> 53.15 | Improved |
| add the attribute prediction term to the criterion (ablation) | zero-shot metric learning model (VGG) · AwA ZSL | accuracy: 75.69 -> 77.32 | Improved |
| add the attribute prediction term to the criterion (ablation) | zero-shot metric learning model (VGG) · CUB-200-2011 ZSL | accuracy: 38.35 -> 43.29 | Improved |
| add the attribute prediction term to the criterion (ablation) | zero-shot metric learning model (VGG) · SUN Attributes ZSL | accuracy: 79.21 -> 84.41 | Improved |
| use VGG-VeryDeep-19 features instead of AlexNet features | zero-shot metric learning model · aPascal&aYahoo ZSL | accuracy: 46.14 -> 53.15 | Improved |
| use VGG-VeryDeep-19 features instead of AlexNet features | zero-shot metric learning model · CUB-200-2011 ZSL | accuracy: 41.98 -> 43.29 | Improved |
| use VGG-VeryDeep-19 features instead of AlexNet features | zero-shot metric learning model · SUN Attributes ZSL | accuracy: 75.48 -> 84.41 | Improved |
| add a few annotated examples of unseen classes and fine-tune (few-shot) | zero-shot metric learning model · ZSL datasets (few-shot) | accuracy | Improved |
Closest papers by meaning
- Preserving Semantic Relations for Zero-Shot Learning (CVPR 2018 · similarity 0.92)
- SimZSL: Zero-Shot Learning Beyond a Pre-defined Semantic Embedding Space (IJCV 2025 · similarity 0.91)
- Zero-Shot Learning on Semantic Class Prototype Graph (TPAMI 2017 · similarity 0.91)
- An Empirical Study and Analysis of Generalized Zero-Shot Learning for Object Recognition in the Wild (ECCV 2016 · similarity 0.90)
- Zero-Shot Learning via Joint Latent Similarity Embedding (CVPR 2016 · similarity 0.90)
- Class Semantic Attribute Perception Guided Zero-Shot Learning (AAAI 2025 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
